A New Focus Strategy for Efficient Dialog Management

@inproceedings{Bao2016ANF,
  title={A New Focus Strategy for Efficient Dialog Management},
  author={Xinqi Bao and Yunfang Wu and Xueqiang Lv},
  booktitle={CCL},
  year={2016}
}
The dialog manager is the most important component for a dialog system, in which the dialog state tracking is crucial to a real-world system. [] Key Method We also implement a partition-based method to deal with the latter problem. Then we combine both strategies to take advantage of their complement property. In our experiment of a real-world application in an image purchase domain, our proposed focus strategy is far faster than both the partition method and the naive algorithm with comparable quality.

References

SHOWING 1-10 OF 23 REFERENCES
Incremental partition recombination for efficient tracking of multiple dialog states
  • J. Williams
  • Computer Science
    2010 IEEE International Conference on Acoustics, Speech and Signal Processing
  • 2010
TLDR
This paper increments partitions by incrementally recombining partitions during the update of spoken dialog states to improve whole-dialog accuracy and views partitions as programmatic objects - an accessible formulation for commercial application developers.
Representing the Reinforcement Learning state in a negotiation dialogue
  • P. Heeman
  • Computer Science
    2009 IEEE Workshop on Automatic Speech Recognition & Understanding
  • 2009
TLDR
This paper explores a task that requires negotiation, in which conversants need to exchange information in order to decide on a good solution, and investigates what information should be included in the system's RL state so that an optimal policy can be learned and so that the state space stays reasonable in size.
User simulation for spoken dialogue systems: learning and evaluation
TLDR
The expected accuracy, expected precision, and expected recall evaluation metrics are proposed as opposed to standard precision and recall used in prior work and discussed why they are more appropriate metrics for evaluating user simulation models compared to their standard counterparts.
A stochastic model of human-machine interaction for learning dialog strategies
TLDR
The experimental results show that it is indeed possible to find a simple criterion, a state space representation, and a simulated user parameterization in order to automatically learn a relatively complex dialog behavior, similar to one that was heuristically designed by several research groups.
A “K Hypotheses + Other” Belief Updating Model
TLDR
A machine-learning based solution for spoken dialog systems that uses a compressed representation of beliefs that tracks up to k hypotheses for each concept at any given time, and trains a generalized linear model to perform the updates.
The Hidden Information State model: A practical framework for POMDP-based spoken dialogue management
Scaling POMDPs for Spoken Dialog Management
  • J. Williams, S. Young
  • Computer Science
    IEEE Transactions on Audio, Speech, and Language Processing
  • 2007
TLDR
A novel POMDP optimization technique-composite summary point-based value iteration (CSPBVI) is presented-which enables optimization to be performed on slot-filling PomDP-based dialog managers of a realistic size and is robust to estimation errors.
...
...